3.3.1. Dependent Variables
The BRFSS contains four HRQOL questions, including general health (GH), physical health (PH), mental health (MH), and activity limitations (AL). This study will analyze each indicator independently.
GH was measured by self-report general health status: “would you say that in general your health is?” PH was measured by “Now thinking about your physical health, which includes physical illness and injury, for how many days during the past 30 days was your physical health not good?” MH was measured by “Now thinking about your mental health, which includes stress, depression, and problems with emotions, for how many days during the past 30 days was your mental health not good?” AL was measured by “During the past 30 days, for about how many days did poor physical or mental health
keep you from doing your usual activities, such as self-care, work, or recreation?” The validity and reliability for public health surveillance for the four HRQOL questions in BRFSS were examined 61,64. Since a large proportion of AL responses were missing
(51.82%), this study did not examine AL. Thus, only the first three indicators, GH, PH and MH, were included.
According to previous research suggestions, GH was divided into two groups: excellent/very good/good and fair/poor. 50,53,65,85,88,115 Although previous research
suggested that PH and MH can be dichotomized into infrequent (0-14 days) and frequent (15-30 days) number of unhealthy days, 50,65,80,88,117 most BRFSS respondents included in
the sample reported no physically (61.42%) or mentally (78.37%) unhealthy days in the past 30 days, skewing the distributions. Thus, PH and MH were categorized into two groups: low unhealthy days (first to third quartiles, 0-4 days (0-74.3%) for PH and 0 days (0-79.4%) for MH) versus high unhealthy days (fourth quartile, 5-30 days (74.3-100%) for PH and 1-30 days (79.4-100%) for MH) (Table 3.4).
Then, HRQOL scores will be aggregated to county-level data based on individual FIPS codes and be used for multilevel modeling. This study applied MPS approach to generate county-level probabilities of having fair/poor GH and high physically/mentally unhealthy days.
Table 3.4: Older adults’ physical health, mental health, and activity limitation distributions before exclusions
Physically unhealthy
days Mentally unhealthy days Activity limitation days
Unweighted Weighted Unweighted Weighted Unweighted Weighted
0 59.87% 58.60% 77.37% 76.10% 1.21% 1.17%
1 2.49% 2.48% 1.95% 2.10% 1.86% 1.90%
2 4.26% 4.44% 3.28% 3.32% 1.28% 1.34%
3 2.73% 2.93% 1.84% 1.93% 0.80% 0.84%
Physically unhealthy days
Mentally unhealthy
days Activity limitation days
Unweighted Weighted Unweighted Weighted Unweighted Weighted
5 2.94% 3.16% 2.08% 2.22% 0.32% 0.31% 6 0.60% 0.65% 0.33% 0.37% 0.79% 0.79% 7 1.72% 1.75% 0.71% 0.76% 0.21% 0.25% 8 0.34% 0.40% 0.20% 0.23% 0.04% 0.04% 9 0.08% 0.09% 0.03% 0.03% 1.46% 1.56% 10 2.45% 2.60% 1.64% 1.76% 0.02% 0.02% 11 0.03% 0.03% 0.01% 0.01% 0.14% 0.17% 12 0.21% 0.26% 0.10% 0.13% 0.01% 0.02% 13 0.02% 0.04% 0.01% 0.01% 0.50% 0.46% 14 1.07% 0.99% 0.29% 0.27% 1.58% 1.69% 15 2.41% 2.64% 1.52% 1.57% 0.03% 0.03% 16 0.05% 0.07% 0.02% 0.02% 0.02% 0.02% 17 0.03% 0.03% 0.01% 0.01% 0.03% 0.03% 18 0.05% 0.05% 0.02% 0.03% 0.01% 0.01% 19 0.01% 0.01% 0.00% 0.01% 0.93% 1.05% 20 1.33% 1.46% 0.79% 0.96% 0.16% 0.16% 21 0.30% 0.30% 0.06% 0.06% 0.02% 0.03% 22 0.03% 0.03% 0.01% 0.01% 0.01% 0.01% 23 0.02% 0.02% 0.01% 0.01% 0.02% 0.02% 24 0.04% 0.03% 0.01% 0.00% 0.43% 0.52% 25 0.62% 0.67% 0.30% 0.38% 0.01% 0.02% 26 0.02% 0.02% 0.01% 0.01% 0.02% 0.02% 27 0.04% 0.04% 0.01% 0.01% 0.09% 0.11% 28 0.18% 0.21% 0.07% 0.09% 0.04% 0.03% 29 0.11% 0.12% 0.05% 0.05% 5.12% 5.20% 30 10.43% 10.20% 3.68% 3.81% 1.62% 1.51% Don’t know/ not sure 3.17% 2.91% 2.01% 1.88% 27.42% 27.76% Refuse d 0.76% 1.07% 0.62% 0.87% 0.39% 0.61% Missin g 0% 0% 0% 0% 51.82% 50.61%
Note: Total unweighted observations before excluded: 313,070; weighted observations before excluded: 86,235,454.
3.3.2. Key Explanatory Variables
Area deprivation index. This study applied Ford and Dzealtowski’s area deprivation index. 26 We chose this particular index because all elements required to calculate the
year from resources): percent of adults unemployed (2011), percent of adults over 25 years with less than a high school education (2008-2012), percent of households under the federally-designated poverty level (2008-2012), percent of households with more than one person per room (2008-2012), percent of female head of household with children (2010), percent of households with public assistance income (2008-2012), median household income (2011), and percent of households with no access to a vehicle (2010).
26 The data for percent of households with no access to a vehicle was obtained from the
2014 Food Environment Atlas Data File, and other data elements from the 2013-2014 AHRF.
We transposed the median household income. Then, each indicator was calculated as a standardized z-score (z =𝑥−𝜇
𝜎 ). Each indicator presents a standard normal
distribution with a mean of 0 and a standard deviation of 1. Then, all indicators are summed into an area deprivation index score, with higher score indicating the area is more deprived. The area deprivation score was divided into two groups based on mean score: low (affluent) versus high (deprived).
3.3.3. Covariates
Covariates contain individual- and county-level characteristics. Individual-level characteristics includes sociodemographic characteristics and health related factors. County-level characteristics were represented by health care resources.
a. Individual-level characteristics (i) Sociodemographic characteristics
Gender sets female as reference group. Age was divided into three categories: 65- 74 years old (reference category), 75-84 years old, or 85 years old and above.
Race/Ethnicity was collapsed into four categories: White (reference category), Black/African American, Hispanic/Latino, or others groups. Others includes Asian, Native Hawaiian/other Pacific Islander, American Indian/Alaska Native, and others.
Educational attainment was divided into two groups: less than high school (<grade 12, reference category) or college and above.
Marital status were also divided into two categories: married and living with spouse and living alone, regardless of marital status (reference category). The second category consisted primarily of divorced, widowed, separated, and never married (single).
Employment status was categorized as three categories: employed, unemployed, or retired (reference category). Employed includes employed for wages, self-employed, and homemakers. Unemployed contains out of work, students, and unable to work.
Annual household income was collapsed into two categories: less than $20,000 (reference category), $20,000 and above, or non-response/missing group.
(ii) Health related factors
Body mass index (BMI) was calculated by weight in kilograms divided by the square of the height in meters (kg/m2) or by weight in pounds divided by the square of the
height in inches and multiply a conversion factor of 703 (lb/in2). BMI was divided into
three categories: optimal weight (18.5-25, reference category), underweight (<18.5), or overweight/obese (>25.0).
Disability was accorded to the question “Are you limited in any way in any activities because of physical, mental, or emotional problems?”, and divided into yes or no (reference category).
day, some days or not at all”), dividing into non-smokers (reference category) or smokers. Current smoking was defined as smoking cigarette every day or some days.
Alcohol consumption was accorded the question “During the past 30 days, how many days per week or per month did you have at least one drink of any alcoholic beverage such as beer, wine, a malt beverage or liquor?” and categorize as non-drinkers (0 days, reference category) or drinkers (at least one day).
Number of chronic conditions were summed up whether they reported having a chronic disease diagnosis, including myocardial infarction, angina or coronary heart disease, stroke, asthma (current having asthma), any type of cancer, chronic obstructive pulmonary disease (COPD)/emphysema/chronic bronchitis, arthritis, depressive disorder, kidney diseases, and diabetes. Those 10 chronic conditions were categorized as 4 groups: 0 (no chronic condition, reference category), 1 (having only 1 chronic conditions), 2 (having any 2 chronic conditions), or 3 and more (having 3 and more chronic conditions). b. County-level characteristics: health care resources
Health care resources were categorized as health facilities and health personnel density. Health facilities factors contain general hospital, county health related centers (including community health center, community mental health center, federal qualified health center, and rural health clinic), and long-term care facility (including long term hospital, skilled nursing facility, and nursing facility). Health personnel density factors contain all primary care providers (including primary care physicians and other primary care providers), and dentists. The data of whole primary care providers gains from the 2014 CHRs, and others are from the 2013-2014 AHRF.
number of each variable divided by total population in the area and multiply 10,000, and presented health resources to population ratios or health personnel densities as the following equation. Then, all county-level factors were divided by mean scores into low (less than mean) or high (larger than mean) groups.
Health care resource to population ratio/health personnel densities
= 𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑒𝑎𝑐ℎ ℎ𝑒𝑎𝑙𝑡ℎ 𝑐𝑎𝑟𝑒 𝑟𝑒𝑠𝑜𝑢𝑟𝑐𝑒
𝑇𝑜𝑡𝑎𝑙 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 ∗ 10,000
Table 3.5: The operational definition of variables
Study variables Operational Definition Attributes
Key explanatory variables
Area deprivation
index In multilevel analysis: 1=low (deprived) 2=high (affluent) Categorical; Numerical (in spatial analysis) Covariates Individual-level Sociodemographic characteristics Gender 1=Male
2=Female (reference category)
Categorical
Age 1=65-74 years (reference category)
2=75-84 years 3=85 years and over
Categorical
Race/Ethnicity 1=White (reference category)
2=Black/African American 3=Hispanic/Latino
4=Others, including Asian/Native Hawaiian or other Pacific Islander, American Indian/Alaska Native, and others
Categorical
Educational attainment
0=Less than high school (<grade 12) (reference category)
1=College and above
Categorical
Marital status 1=Married/A member of an unmarried (reference
category)
2=Divorced/Widowed/Separated/Never married
Categorical
Employment status 1=Employed
2=Unemployed
3=Retired (reference category)
Categorical
Study variables Operational Definition Attributes
income 2=$20,000 and above
3=Non-response, don’t know/not sure, or missing Health related
factors
Body Mass Index 1=Optimal weight (reference category)
2=Underweight 3=Overweight/obese
Categorical
Disability 0=No (reference category)
1=Yes Categorical
Smoking 0=Non-smoking (reference category)
1=Current smoking (every day or some days) Categorical
Alcohol consumption
0=No drinking (reference category) 1=Drinking
Categorical Number of chronic
conditions
Summed up whether they reported having a chronic disease diagnosis: myocardial infarction, angina or coronary heart disease, stroke, asthma (current having asthma), any type of cancer, chronic obstructive pulmonary disease
(COPD)/emphysema/chronic bronchitis, arthritis, depressive disorder, kidney diseases, and diabetes. Grouping up:
0=0 chronic condition (reference category) 1=1 chronic condition 2=2 chronic condition 3=≥3 chronic condition Categorical Community-level Health facilities
factors Including: 1. General hospital to population ratio 2. County health related centers
a. Community health center
b. Community mental health center c. Federal qualified health center d. Rural health clinic
3. Long-term care facility to population ratio a. Long term hospital
b. Skilled nursing facility c. Nursing facility
Dividing by mean scores: 1=low (< mean) 2=high (≥ mean) Categorical Health personnel density factors Including:
1. All primary care providers density a. Primary care physicians b. Other primary care providers 2. Dentists density
Study variables Operational Definition Attributes Dividing by mean scores:
1=low (< mean) 2=high (≥ mean) Dependent Variables
General health In multilevel analysis:
0=excellent/very good/good 1=fair/poor
In spatial analysis: calculating by MPS approach
Categorical; Numerical
Physical health In multilevel analysis:
0=low unhealthy days (first to third quartiles) 1=high unhealthy days (fourth quartiles)
In spatial analysis: calculating by MPS approach
Categorical; Numerical
Mental health In multilevel analysis:
0=low unhealthy days (first to third quartiles) 1=high unhealthy days (fourth quartiles)
In spatial analysis: calculating by MPS approach
Categorical; Numerical